Jixiang Lei
- Telephone number
- office: +43 316 873 - 30094
- Position
- PhD student
- jixiang.lei@tugraz.at
- Other Profiles
Jixiang Lei is a PhD student at the Institute of Signal Processing and Speech Communication, Graz University of Technology, under the supervision of Univ.-Prof. Dr. Franz Pernkopf. His research focuses on robust and adaptive machine learning, particularly in the areas of test-time adaptation, multimodal learning, and deep learning under distribution shifts. He holds a Master’s degree in Artificial Intelligence from Johannes Kepler University Linz, where he specialized in robotics and semantic scene understanding, and a Master’s degree in Aerospace Engineering from the University of Stuttgart. Prior to his PhD, he worked in both academic and industrial research environments, contributing to projects in predictive maintenance, condition monitoring, and computer vision. His recent work has been presented at leading conferences such as ICML Workshop and IJCNN, and he is particularly interested in the translation of ML theory into robust, real-world applications in safety-critical domains.
Publications
- Conference paper Lei J. & Pernkopf F. (2025) Two-Level Test-Time Adaptation in Multimodal Learning. in 2025 International Joint Conference on Neural Networks, IJCNN 2025. [more info]
- Conference paper Lei J. & Pernkopf F. (2025) HR-OTTA: Robust Online Test-Time Adaptation with Hybrid Fine-Tuning. in 2025 International Joint Conference on Neural Networks, IJCNN 2025. [more info] [doi]
- Conference paper Lei J. & Pernkopf F. (2025) PromptCAL: Entropy-Calibrated and Prompt-Tunes Test Time Adaptation for Semantic Segmentation. in 2025 IEEE International Conference on Systems, Man, and Cybernetics, SMC 2025. [more info]
- Conference paper Lei J. & Pernkopf F. (2025) GaPaTTA: Gaussian Entropy-Guided Prompt Placement for Continual Test-Time Adaptation in Semantic Segmentation. in 17th Asian Conference on Machine Learning, ACML 2025. [more info]
- Journal article Lei J. & Pernkopf F. (2025) GaPaTTA: Gaussian Entropy-Guided Prompt Placement for Continual Test-Time Adaptation in Semantic Segmentation. in Machine Learning. [more info]
- Conference paper Lei J. & Pernkopf F. (2024) Two-Level Test-Time Adaptation in Multimodal Learning. in ICML 2024 Workshop on Foundation Models in the Wild, ICML 2024 FM-Wild. [more info]